Structural analyses of DNA recognition by the AML1/Runx-1 Runt domain and its allosteric control by CBFbeta

The core binding factor (CBF) heterodimeric transcription factors comprised of AML/CBFA/PEBP2alpha/Runx and CBFbeta/PEBP2beta subunits are essential for differentiation of hematopoietic and bone cells, and their mutation is intimately related to the development of acute leukemias and cleidocranial dysplasia. Here, we present the crystal structures of the AML1/Runx-1/CBFalpha(Runt domain)-CBFbeta(core domain)-C/EBPbeta(bZip)-DNA, AML1/Runx-1/CBFalpha(Runt domain)-C/EBPbeta(bZip)-DNA, and AML1/Runx-1/CBFalpha(Runt domain)-DNA complexes. The hydrogen bonding network formed among CBFalpha(Runt domain) and CBFbeta, and CBFalpha(Runt domain) and DNA revealed the allosteric regulation mechanism of CBFalpha(Runt domain)-DNA binding by CBFbeta. The point mutations of CBFalpha related to the aforementioned diseases were also mapped and their effect on DNA binding is discussed.

Crystal structure of abrin-a at 2.14 A

The crystal structure of abrin-a, a type II ribosome-inactivating protein from the seeds of Abrus precatorius, has been determined from a novel crystalline form by the molecular replacement method using the coordinates of ricin. The structure has been refined at 2.14 A to a R-factor of 18.9%. The root-mean-square deviations of bond lengths and angles from the standard values are 0.013 A and 1.82 degrees, respectively. The overall protein folding is similar to that of ricin, but there are differences in the secondary structure, mostly of the A-chain. Several parts of the molecular surface differ significantly; some of them are quite near the active site cleft, and probably influence ribosome recognition. The positions of invariant active site residues remain the same, except the position of Tyr74. Two water molecules of hydrogen-bonded active site residues have been located in the active site cleft. Both of them may be responsible for hydrolyzing the N-C glycosidic bond. The current abrin-a structure is lactose free; this is probably essential for abrin-a crystallization. The B-chain is a glycoprotein, and the positions of several sugar residues of two sugar chains linked to earlier predicted glycosylation sites were determined. One of the sugar chains is a bridge between two neighboring molecules, since one of its mannose residues is connected to the galactose binding site of the neighboring molecule. Another sugar chain covers the surface of the B-chain.

Crystal structure of methionine aminopeptidase from hyperthermophile, Pyrococcus furiosus

The structure of methionine aminopeptidase from hyperthermophile Pyrococcus furiosus (PfMAP) with an optimal growth temperature of 100 degreesC was determined by the multiple isomorphous replacement method and refined in three different crystal forms, one monoclinic and two hexagonal, at resolutions of 2.8, 2.9, and 3.5 A. The resolution of the monoclinic crystal form was extended to 1.75 A by water-mediated transformation to a low-humidity form, and the obtained diffraction data used for high-resolution structure refinement. This is the first description of a eukaryotic type methionine aminopeptidase structure. The PfMAP molecule is composed of two domains, a catalytic domain and an insertion domain, connected via two antiparallel beta-strands. The catalytic domain, which possesses an internal 2-fold symmetry and contains two cobalt ions in the active site, resembles the structure of a prokaryotic type MAP from Escherichia coli (EcMAP), while the structure of the insertion domain containing three helices has a novel fold and accounts for a major difference between the eukaryotic and prokaryotic types of methionine aminopeptidase. Analysis of the PfMAP structure in comparison with EcMAP and other mesophile proteins reveals several factors which may contribute to the hyperthermostability of PfMAP: (1) a significantly high number of hydrogen bonds and ion-pairs between side-chains of oppositely charged residues involved in the stabilization of helices; (2) an increased number of hydrogen bonds between the positively charged side-chain and neutral oxygen; (3) a larger number of buried water molecules involved in crosslinking the backbone atoms of sequentially separate segments; (4) stabilization of two antiparallel beta-strands connecting the two domains of the molecule by proline residues; (5) shortening of N and C-terminal tails and stabilization of the loop c3E by deletion of three residues.

High-resolution crystal structures of two polymorphs of cytochrome c' from the purple phototrophic bacterium rhodobacter capsulatus

The structures of two polymorphs of cytochrome c' from Rhodobacter capsulatus (RCCP) strain M110 have been determined by the molecular replacement method. Iron anomalous scattering data were used to confirm the molecular replacement solution. The structures were refined at 1.72 angstrom and 2.0 angstrom resolution to R-values of 15.0% and 16.3%, respectively. The RCCP molecule is a dimer and each of the identical 129 residue subunits folds as a four-helical bundle with a covalently bound heme group in the center. This structural motif resembles that of cytochromes c' reported from Rhodospirillum molischianum (RMCP), Rhodospirillum rubrum (RRCP), Chromatium vinosum (CVCP), Achromobacter xyloseoxidans (AXCP) and Alcaligenes denitrificans (ADCP). However, the architecture of the RCCP dimer, that is, the mode of association of subunits, differs substantially from that of the other cytochromes c'. In the RCCP dimer, the subunits are roughly parallel with each other and only helix B of each subunit participates in formation of the dimer interface. Measurement of the solvent-accessible surface area indicates that the dimer interface is smaller in RCCP than in the other cytochromes c'. In RMCP, CVCP, RRCP, AXCP and ADCP the subunits cross each other to form an X shape, and two helices, A and B, of each subunit interact across the dimer interface. These results are consistent with hydrodynamic measurements, which show that there is an equilibrium between monomers and dimer in RCCP, whereas the dimer is the predominant form in the other cytochromes c' for which structures have been determined. Structural comparison of the six cytochromes c' reveal that they can be divided into two groups. In group 1 cytochromes c', CVCP and RCCP, the amino acid sequences and the folding of subunits are arranged in such a way as to allow the formation of a deep channel between helices B and C with direct solvent accessibility to the heme sixth ligand position. There is no such channel in group 2 cytochromes c', RMCP, RRCP, AXCP and ADCP. This may account, in part, for the differences in carbon monoxide binding.

Structural basis of Ets1 activation by Runx1

Runx1 is required for definitive hematopoiesis and is well-known for its frequent chromosomal translocations and point mutations in leukemia. Runx1 regulates a variety of genes via Ets1 activation on an Ets1•Runx1 composite DNA sequence. The structural basis of such regulation remains unresolved. To address this problem, we determined the crystal structure of the ternary complex containing Runx1_1-242 and Ets1_296-441 bound to T cell receptor alpha (TCRα) enhancer DNA. In the crystal, an Ets1-interacting domain of Runx1 is bound to the Ets1 DNA-binding domain and displaced an entire autoinhibitory module of Ets1, revealing a novel mechanism of Ets1 activation. The DNA binding and transcriptional studies with a variety of structure-guided Runx1 mutants confirmed a critical role of direct Ets1•Runx1 interaction in Ets1 activation. More importantly, the discovered mechanism provides a plausible explanation for how the Ets1•Runx1 interaction effectively activates not only a wild-type Ets1, but also a highly inhibited phosphorylated form of Ets1.

Concerted movement of side chains in the haem vicinity observed on ligand binding in cytochrome c' from rhodobacter capsulatus

We have determined the structure of n-butylisocyanide-bound Rhodobacter capsulatus cytochrome c'. This is the first example of a ligand-bound structure of a class IIa cytochrome c. Compared with the structure of native cytochrome c', there are significant conformational changes of amino acid residues in the haem vicinity, accompanied by a rearrangement of the hydrogen bonding pattern. The results suggest that rearrangements resulting from ligand binding could drive dimer dissociation in some species and also that the haem propionate may participate in proton transfer.

Nucleotide selectivity defect and mutator phenotype conferred by a colon cancer-associated DNA polymerase δ mutation in human cells

Mutations in the POLD1 and POLE genes encoding DNA polymerases δ (Polδ) and ɛ (Polɛ) cause hereditary colorectal cancer (CRC) and have been found in many sporadic colorectal and endometrial tumors. Much attention has been focused on POLE exonuclease domain mutations, which occur frequently in hypermutated DNA mismatch repair (MMR)-proficient tumors and appear to be responsible for the bulk of genomic instability in these tumors. In contrast, somatic POLD1 mutations are seen less frequently and typically occur in MMR-deficient tumors. Their functional significance is often unclear. Here we demonstrate that expression of the cancer-associated POLD1-R689W allele is strongly mutagenic in human cells. The mutation rate increased synergistically when the POLD1-R689W expression was combined with a MMR defect, indicating that the mutator effect of POLD1-R689W results from a high rate of replication errors. Purified human Polδ-R689W has normal exonuclease activity, but the nucleotide selectivity of the enzyme is severely impaired, providing a mechanistic explanation for the increased mutation rate in the POLD1-R689W-expressing cells. The vast majority of mutations induced by the POLD1-R689W are GC→︀TA transversions and GC→︀AT transitions, with transversions showing a strong strand bias and a remarkable preference for polypurine/polypyrimidine sequences. The mutational specificity of the Polδ variant matches that of the hypermutated CRC cell line, HCT15, in which this variant was first identified. The results provide compelling evidence for the pathogenic role of the POLD1-R689W mutation in the development of the human tumor and emphasize the need to experimentally determine the significance of Polδ variants present in sporadic tumors.

High-resolution crystals of methionine aminopeptidase from Pyrococcus furiosus obtained by water-mediated transformation

The monoclinic crystal form of methionine amino-peptidase from Pyrococcus furiosus (MAP-Pfu) has been crystallized from four different conditions. Native crystals belong to space group P2(1) with typical unit-cell dimensions a = 53.4, b = 85.1, c = 72.7 A, beta = 107.7 degrees and diffract to 2.9-4.5 A resolution. However, there is a problem of nonisomorphism among the crystals. Water-mediated transformation to low-humidity form occurs by reduction of the relative humidity of crystal environment to 79%. The unit-cell dimensions of transformed crystals are a = 51.9, b = 83.3, c = 70.3 A, beta = 105.9 degrees, and the calculated solvent content is 3.9% less than in original crystals. Transformation to low-humidity form is accompanied by 1.7 times reduction of overall temperature factors, extension of diffraction resolution up to 1.75 A, without change or reduction of crystal mosaicity, and improvement in stability to X-ray radiation. The water-mediated transformation also appears to relieve the problem of nonisomorphism among the original MAP-Pfu crystals.

Crystals of ternary protein-DNA complexes composed of DNA-binding domains of c-Myb or v-Myb, C/EBPalpha or C/EBPbeta and tom-1A promoter fragment

c-Myb and the C/EBP family are transcriptional regulatory factors that act in concert to regulate the expression of myeloid-specific genes. v-Myb encoded by avian myeloblastosis virus (AMV) is a mutated form of c-Myb that contains point mutations which disrupt the cooperation with C/EBPs. To understand the mechanism of the transcriptional synergy between c-Myb and C/EBPs and the effect of the v-Myb mutations on that synergy, knowledge based on their three-dimensional structures is essential. Crystals of ternary complexes, in which various combinations of the DNA-binding domains of c-Myb or v-Myb and C/EBPalpha or C/EBPbeta are bound to a DNA fragment from tom-1A promoter, were obtained by the vapour-diffusion method. Complete diffraction data sets were obtained from each native crystal and two types of iodine-derivative crystals. A three-wavelength MAD data set was also obtained from a bromine-derivative crystal.

Crystallization and preliminary X-ray analyses of quaternary, ternary and binary protein-DNA complexes with involvement of AML1/Runx-1/CBFalpha Runt domain, CBFbeta and the C/EBPbeta bZip region

Three types of protein-DNA complexes, AML1/Runx-1/CBFalpha(Runt)-CBFbeta-C/EBPbeta(bZip)-DNA (CBFalpha-beta-C/EBPbeta-DNA), AML1/Runx-1/CBFalpha(Runt)-C/EBPbeta(bZip)-DNA (CBFalpha-C/EBPbeta-DNA) and AML1/Runx-1/CBFalpha(Runt)-DNA (CBFalpha-DNA), were crystallized. The crystals were all orthorhombic and belonged to space groups C222(1), P2(1)2(1)2 and P2(1)2(1)2(1), respectively. The resolutions of CBFalpha-beta-C/EBPbeta-DNA and CBFalpha-C/EBPbeta-DNA crystals were both 3 A, while that of the CBFalpha-DNA crystal was 2.65 A. Complete data sets were collected for all of the native crystals, along with MAD and MIR data sets for CBFalpha-beta-C/EBPbeta-DNA. The heavy-atom site was determined using MAD data for a gold derivative of CBFalpha-beta-C/EBPbeta-DNA.

Structure of Cytochromec' fromRhodobacter capsulatusStrain St Louis: An Unusual Molecular Association Induced by Bridging Zn Ions

Rhodobacter capsulatus strain St Louis cytochrome c' (RCCP-SL) has been crystallized and the structure solved by molecular replacement. It was refined at 2.1 A resolution to an R value of 18.4%, and compared with Rhodobacter capsulatus strain M110 cytochrome c' (RCCP-M110). Although these two proteins are very similar in sequence and structure, the intermolecular interaction is largely different. In RCCP-M110, the molecules dimerize through interaction of helix B to form an antiparallel arrangement. When crystallized in the presence of Zn ions, molecules of RCCP-SL were found to be arranged as linear polymers connected by the bridging Zn ions. The changes in conformation of the side chains induced by binding of the Zn ions, by the substitution of Glu90 for Asp90, and by the different arrangement of the molecules, are discussed in detail.

Crystallization and preliminary X-ray analysis of the C/EBPbeta C-terminal region in complex with DNA

The C-terminal fragment (residues 259-345) of human C/EBPbeta, a basic region leucine zipper transcriptional regulatory factor which includes the minimal DNA-binding domain, was crystallized in complex with a 16 bp DNA fragment from the tom-1 promoter. The crystals were in the form of a parallelepiped belonging to space group C222(1), had unit-cell parameters a = 100.7 (2), b = 113.5 (1), c = 74.4 (1) A and diffracted to a resolution of 2.1 A. Moreover, truncation of nine residues from the C-terminus not conserved among C/EBP family members yielded isomorphous crystals that diffracted to a resolution of 1.8 A or better. Truncation of 14 residues from the N-terminus of the C-terminal fragment produced well shaped crystals in the form of hexagonal bipyramids, however; unfortunately, they were unstable and diffracted poorly.

Crystallization and preliminary X-ray analysis of wild-type and V103L mutant Myb R2 DNA-binding domain

The R2 subdomain of the mouse c-Myb DNA-binding domain and its V103L mutant have been crystallized by the vapour-diffusion method using highly concentrated sodium citrate at pH 6.8 as a precipitant. Using ammonium sulfate as precipitant in MES buffer only produced crystals for the mutant R2. All crystals are isomorphous and belong to space group P212121. The unit-cell dimensions for wild-type R2 crystals grown from sodium citrate precipitant are a = 28.83, b = 40.18, c = 49.23 A. Crystals contain one R2 molecule per asymmetric unit. They are stable during 3 d exposure to X-rays and diffract to 1.37-1.45 A resolution.

Crystallization and preliminary X-ray analysis of methionine aminopeptidase from the hyperthermophilic bacterium Pyrococcus furiosus

Methionine aminopeptidase (MAP) from Pyrococcus furiosus (Pfu) has been crystallized in four different forms (A, B, C and D). Form A crystals belong to space group P2(1) with unit-cell dimensions a = 54.18, b = 85.72, c = 72.84 A, beta = 108.34 degrees. Forms B, C and D belong to space group P6(2(4)) with unit-cell dimensions a = 139.1, c = 63.7 A for form B, a = 198.6, c = 243.8 A for form C, and a = 111.0, c = 125.0 A for form D. Forms A and D diffract to 2.9 A, form B diffracts to 3.5 A, and form C crystals diffract to 4.5 A. Form A contains two molecules of MAP-Pfu per asymmetric unit. The binuclear metal center positions and a non-crystallographic twofold symmetry matrix has been determined for the form A crystals.

A new crystal form of abrin-a from the seeds of Abrus precatorius

A new crystal form of abrin-a from the seeds of Abrus precatorius has been obtained by vapor diffusion method. The abrin-a crystals belong to monoclinic space group P2(1) with cell dimensions a = 84.58 A, b = 73.07 A, c = 48.23 A, beta = 96.20 degrees. An asymmetric unit contains one protein molecule of molecular weight 65 kDa and has a solvent content of approximately 46%.